ACR Methodology for the Conversion of High-Bleed Pneumatic Controllers in Oil and Natural Gas...

8/2/2019 ACR Methodology for the Conversion of High-Bleed Pneumatic Controllers in Oil and Natural Gas Systems v1.1

1/33

AmericanCarbonRegistry

American Carbon RegistryTrusted solutions for the carbon market

Anonprofitenterpriseof

2010 All Rights Reserved

EmissionReductionMeasurementand

MonitoringMethodologyforthe

ConversionofHighBleedPneumaticControllersinOilandNaturalGasSystems

MethodologydevelopedbyVerdeoGroup,Inc.andDevonEnergyCorporationVersion1.0publishedMarch2010;revisedversion1.1publishedDecember2010


2/33


3/33

Emission Reduction Measurement and Monitoring Methodology for theConversion of High-Bleed Pneumatic Controllers in Oil and Natural Gas Systems

Page 2

1.0 BACKGROUND AND APPLICABILITY

1.1 BackgroundonPneumatic

Controllersin

the

Oil

and

Natural

GasIndustry

Pneumatic controllers use a pressurized gas,

typically natural gas for applications in the oil

and gas production industry, to regulate

process variables such as pressure, flow rate

and liquid level. Pneumatically operated

equipment is used in the oil and gas industry

because electricity is not readily available at

remote production sites.

Most pneumatic instruments and controllers in

the natural gas industry are powered by natural

gas, and these controllers are designed to

discharge methane to the atmosphere as a part

of normal operations. Pneumatic controllers can

be designed to bleed at both high and low-

bleed-rates. Before 1990, all controllers were

designed with generally high-bleed rates.

Increasing awareness of the extent of wasted

resources and potential environmental hazards

resulting from higher bleed rates led to the

development and introduction of low-bleed

pneumatic controllers in the early 1990s.1 It has

now become standard practice to use low-bleed

pneumatic controls in new constructionin the oil

and gas industry. Despite the existence of low-

bleed technology as well as retrofit solutions

such as WellMarks Mizer Pilot Valve,

conversions of existing high-bleed controllers

1The

EPA

defines

high

bleed

controllers

as

controllers

thatbleedinexcessof6cubicfeetperhour,andalow

bleedpneumaticcontrollerisdefinedasacontrollerthat

bleedsatarateoflessthan6cubicfeetperhour.

USEnvironmentalProductionAgency,LessonsLearned

fromNaturalGasSTARPartners:OptionsforReducing

MethaneEmissionsfromPneumaticDevicesinthe

NaturalGasIndustry,EPA430B03004,Washington,

DC,July2003.

are uncommon. Numerous sources estimate

that hundreds of thousands of high-bleedpneumatics still remain in service throughout

the oil and gas production segment. As a result,

high-bleed pneumatic controllers are among the

largest sources of vented methane by

equipment type in the domestic oil and natural

gas industry.2

The objective of this methodology is to provide

a method for quantifying baseline emissions

from high-bleed pneumatic devices for the

purpose of generating tradable GHG emissionreduction offsets to incentivize the development

of a voluntary, pre-compliance offsets market in

the U.S., as well as the rapid adoption and

implementation of a technological solution for

an important GHG emissions source.

1.2 Applicability

This Baseline Monitoring Methodology for the

Conversion of High-Bleed Pneumatic

Controllers in Oil and Natural Gas Systems[Pneumatic Conversion Methodology]describes the baseline measurement and post-

project monitoring procedures for the retrofit or

replacement of high-bleed pneumatic

controllers with low-bleed alternatives for the

purpose of an emission reductions project

where emission reductions can be registered.

This methodology is applicable only for the

conversion of high-bleed pneumatic controllers

to low-bleed pneumatic controllers, whereoperating conditions and requirements permit

such conversions. In some operational

applications e.g., the control of very large

valves that require fast or precise process

2InventoryofU.S.GreenhouseGasEmissionsandSinks

19902006,USEPA,April,2008.


4/33


Page 3

changes, high-bleed controllers should not be

replaced with low-bleed controllers. In these

cases project proponents should investigate the

use of new fast-acting devices with lower bleed

rates.

This methodology is applicable to continuous

bleed pneumatic controllers in level, flow rate,

temperature or pressure control applications.

Constraints on the applicability of the

methodology are as follows:

This methodology is applicable to the

conversion of both snap-acting and throttle

acting high-bleed pneumatic controllers to

low-bleed pneumatic controllers. This methodology is not applicable to

projects which deal with the conversion of

high-bleed pneumatic controllers in sectors

outside of oil and natural gas systems.

This methodology is only for the conversion

of existing high-bleed rate pneumatic

controllers, and is not applicable to newly

built oil and gas production facilities that are

installing low-bleed pneumatic controllers.

Projects using this methodology will have acrediting period of ten (10) years.

1.3 PeriodicReviewsandRevisions

The American Carbon Registry (ACR) may

require revisions to this methodology to ensure

that monitoring, reporting, and verification

systems adequately reflect changes in the

projects activities. An annual attestation and a

verification statement, which will include

documentation of the findings of a third party

verifier applying audit sampling methods,

should be submitted by the project entity to

ACR. The ACR will then:

Review the attestation and verification

statement and notify the project entity of any

required adjustments or corrections to these

documents.

Register verified emission reductions.

This methodology may also be periodically

updated to reflect regulatory changes, emission

factor revisions, or expanded applicability

criteria including new models of eligible high-

bleed controllers or additional oil and gas

industry segments. Before beginning a project,

the project proponent should ensure that they

are using the latest version of the methodology.


5/33


Page 4

2.0 BASELINE DETERMINATION

2.1 Project

Assessment

Boundaries

andEmissionsSources

The project boundaries will be confined to all

conversions implemented by a single project

proponent in a continuous time frame or in

contiguous phases.

Because the project implementation, key

project data collection and storage, calculation

of the project emissions, and total emission

reductions will be centrally managed and stored

by the project proponent, then it is logical thatthe GHG emission reduction project

assessment boundaries will be the universe of

conversions undertaken by the project

proponent. However, due to unique operating

conditions (i.e. pressure or methane

concentration) which could affect emission

rates, certain baseline and project emissions

will need to be calculated at a basin level in the

production sector and at a facility or pipeline

level in the transmission and distributionsectors, as described below.

Baseline emissions for the project will be

determined through site-specific sampling. The

project proponent can establish a manufacturer-

specific emission factor for its own population of

pneumatic controllers per Sections 3.1.1 & 4.4

and Appendix C below.

Project emission factors will be extrapolated

from a series of representative sample

measurements of the converted project

population. Further information on the

measurement and calculation of emissions can

be found in the project and baseline emissions

calculations as per Sections 3.2 and 4.4 and

Appendix C below.

There are no emissions from construction or

combustion as a part of the pneumatic retrofit

project. Emissions in the project baseline and

post-retrofit scenario consist solely of the GHG

fraction of natural gas emissions from the

pneumatic controllers. These gases, and their

status (included or excluded) within the project

are outlined in Table 2.1.

Table2.1:BaselineandProjectEmissions

Gas DescriptionStatus:

(included, excluded)

CH4

Methane is a major constituent of natural gas. Themethane composition of natural gas tends to bebetween 70% and 90%, but varies depending onthe location of the production, transmission or

distribution facility. The methane fraction of thegas is an important component in the emissionreductions calculation. Facilities will be required topresent information from a gas chromatograph orcomparable test.

Methane is included as a project andbaseline emission.


6/33


Page 5

Gas DescriptionStatus:

(included, excluded)

CO2 CO2 constitutes about 5% of most natural gasemissions.

Excluded. While the retrofit ofcontrollers will result in the reductionof CO2 released bleeding to the

atmosphere, no credit will be claimedfor this reduction in order to ensurethat the project emission reductionclaims are conservative.

NMHCs

Non-methane hydrocarbon components of naturalgas can include ethane, propane, butane,pentane, or any other non-methane gasses (e.g.nitrogen, helium, and hydrogen sulfide) found inthe natural gas.

Excluded

In certain applications, controllers may reach

the end of their useful life during the crediting

period for the project, such as in cases where a

controller is worn down more quickly if the

application uses corrosive gas. To determine if

any controller included in the project would

normally have been replaced with alow-bleed

alternative during the crediting period for a

reason other than the project activity, the

project proponent should provide the following

information:

The project proponent should describe

current practice for routine refurbishment of

controllers and should provide to the verifier

the standard operating procedure, if any, for

routine refurbishment of pneumatic

controllers, including replacement

specifications published by the controllers

manufacturer if available.

Any controllers which would have been so

replaced during the crediting period should

be identified, the expected date of suchreplacement should be stated, and no

emission reductions credited for that

controller after such date.

2.2 Baseline

Description

The baseline scenario is the continued use of

high-bleed pneumatic controllers. Project

entities must demonstrate that the pneumatic

conversion project is not common practice per

the Practice-based Performance Standard

defined below.

2.3 AdditionalityAssessment

Emission reductions from the project must be

additional, or deemed not to occur in the

business-as-usual scenario. Assessment of the

additionality of a project will be made based on

passing the two tests cited below. These two

tests require the project proponent to

demonstrate that the project activity is surplus

to regulations and reduces emissions below the

level established, through the practice-based

performance standard as defined below, to

represent common practice or business-as-

usual for the retrofit of high-bleed pneumatic

controllers with low or no-bleed alternatives.

Project proponents utilizing this methodology

should consult the latest version of the ACR

Standard, which may be updated from time to

time. At the time of the drafting of this

methodology, the two additionality tests include:


7/33


Page 6

1. Regulatory Surplus Test, and

2. Practice-based Performance Standard

Further guidance on these tests is given below:

TEST1:

Regulatory

Surplus

In order to pass the regulatory surplus test, a

project must not be mandated by existing laws,

regulations, statutes, legal rulings, or other

regulatory frameworks in effect now, or as of

the project start date, that directly or indirectly

affect the credited GHG emissions associated

with a project.

The project proponent must demonstrate that

there is no existing regulation that mandates

the project or effectively requires the GHG

emission reductions associated with the retrofit

of high-bleed pneumatic controllers with low or

no-bleed alternatives.

TEST2:PracticebasedPerformanceStandard

An assessment of the market penetration of

high- to low-bleed retrofits, based on national

and regional market information from various

government agencies (federal, state, local),

equipment vendors and trade associations,demonstrates that market conditions have

resulted in a common practice conversion rate

for high-bleed pneumatic devices of less than

10%. Because the conversion of high-bleed

pneumatic controllers with low-bleed

alternatives is not common practice in the oil

and gas industry, retrofit projects using this

methodology are deemed beyond business as

usual and therefore additional.

Projects which meet the eligibility criteria for thismethodology can use the performance standard

to demonstrate additionality without providing

additional implementation barrier analysis.

Projects that are certified under this version of

the methodology do not need to reassess

additionality with each verification during their

10-year crediting period. However, the

following common practice assessment and the

applicability of the practice-based performance

standard will be reassessed periodically after

significant changes to the market, or, at a

minimum, every 10 years. Future commonpractice assessments should differentiate

between retrofits that occurred as emission

reduction projects and business-as-usual.

ACR reserves the right to review the common

practice assessment as necessary to ensure

additionality of future projects. All GHG Project

Plans for new projects, and all applications for

crediting period renewal on existing projects,

shall apply the regulatory surplus and practice-

based performance standard tests in the latestapproved revision of this methodology in effect

at the time of GHG Project Plan submission or

application for crediting period renewal.

Common Practice Assessment for Conversion

of High-Bleed Pneumatic Controllers: Because

pneumatic controllers are numerous and

dispersed throughout the oil and gas

production, gathering, and transmission

segments, and no organization has undertaken

a comprehensive inventory, it is impossible tohave a precise estimate of the number of

pneumatic controllers or their aggregate

emissions. However, the U.S. EPA, for its

annual Inventory on Greenhouse Gas

Emissions and Sinks, routinely performs top-

down estimations of pneumatic populations and

emissions using available industry activity data

and emission factors. Based on this EPA data

and further analysis of population size and

penetration rates it is estimated that less than

10% of the pneumatic high-bleed controllerpopulation has been replaced with low-bleed

pneumatic devices.

According to the EPAs study, emissions from

pneumatic controllers in the oil and gas sector

account for 48 billion cubic feet of natural gas


8/33


Page 7

emissions per annum.3 In the same study, the

EPA also estimated that there were

approximately 498,000 pneumatic controllers in

the oil and gas sectors. This number includes

high and low-bleed pneumatic controls.Additionally, using the EPAs estimate of total

emissions and estimates of average emissions

from each high-bleed controller (~140,000

cf/year) and low-bleed controllers (~8,000

cf/year) the total number of pneumatic

controllers is estimated at closer to 525,650.

Further, the U.S. EPA undertook a study in

2002 that estimated that the ratio of high-bleed

to low-bleed pneumatic controls in the oil and

gas sector was 66% to 34%.4

The study alsoimplied that the number of existing high-bleed

pneumatics was diminishing at a very low rate;

this is essentially because few replacements

are occurring. These ratios, along with total

population figures, imply that the total number

of remaining high-bleed population in the United

States is currently between 328,680 and

346,930 controllers.

Like the total number of pneumatic controllers,

the number of retrofits/replacements that haveoccurred is equally difficult to determine

precisely. However, data from the EPA and

industry vendors and equipment suppliers

suggest that the market penetration rate for

high-bleed conversion in the production sector

remains low.

Since 1993, the EPA has run a voluntary

program called Gas STAR to educate natural

gas companies about opportunities to reduce

emissions from their operations. There are over

3USEnvironmentalProductionAgency(EPA),Lessons

LearnedfromNaturalGasSTARPartners:Optionsfor

ReducingMethaneEmissionsfromPneumaticDevicesin

theNaturalGasIndustry,EPA430B03004,

Washington,DC,July2003.4Moreinformationonthisstudymaybeaccessedby

contactingtheEPAsNaturalGasSTARprogram.

100 Gas STAR members, of which more than

30 are natural gas production companies,

including 21 of the 25 largest domestic

producers. These members submit annual

reports to Gas STAR describing the annualemission reduction projects completed during

the year. According to the EPA data, members

have reported replacing approximately 34,000

high-bleed devices since 1990(this figure

includes both transmission and distribution

companies as well as production companies).

Combining the above estimate of 328,680

346,930 high-bleed controllers (in the oil and

gas sectors) with the EPA figure that 34,000

high-bleed conversions have occurred (in bothproduction and transmission sectors since

1990) would suggest that the market

penetration of low-bleed conversions in the

industry is below 10%.

This penetration rate is further supported by

discussions with vendors. Though vendors do

not track whether their low-bleed devices are

sold for conversion projects or new installations,

anecdotal evidence suggests that such

replacement projects are uncommon. Vendorsurveys indicate that approximately 20,000

devices have been retrofitted since 2000,

implying a market penetration rate well below

10% and natural conversion rate of less than

1%.

In summary, despite superior technology having

been available in the market for 20 years, a

very large number of high-bleed pneumatic

controllers still exist. The rate of conversion of

these controllers, either through the retirement

of the equipment on which they operate, or

through retrofits, is extremely slow. Anecdotal

evidence from industry professionals and field

operators indicates that there are a number of

reasons for this slow natural rate of conversion,

including lack of understanding of the new

technology among field operators, a pervasive


9/33


Page 8

if it aint broke, dont fix it attitude, scarcity of

capital and human resources for non-core

projects, and corporate budgeting practices that

do not account for lifecycle efficiency costs.

Additionally, operators are reluctant to act on

emission reduction activities before there is a

clear path forward on U.S. GHG regulations

and law and are adopting a wait and see

attitude.


10/33


Page 9

3.0 QUANTIFICATION METHODOLOGY

3.1 Baseline

Emissions

Quantification of project emission reductions

will require baseline emissions and project

emissions, which are calculated from emission

factors derived through statistical sampling and

analysis of the project proponents population of

pneumatic controllers as follows:

Within the project boundary, an emission factor

for calculating baseline emissions will be

developed from site-specific measurements

using procedures outlined in Appendix C. Arepresentative sample, of not less than 30

direct measurements, must be taken to

estimate the average emissions of each

manufacturers controllers to be converted in

the project. As there is a statistically significant

difference between the average emissions of

different manufacturers controllers (Appendix

A) sampling and emission factors should be

manufacturer-specific. Using the data from the

sample of measurements the lower bound 95%

confidence interval for the average emissions

will be used as the emission factor to calculate

the baseline. By using the lower bound of the

confidence level, the uncertainty around

sampling is automatically discounted, thereby

providing a conservative result.

All measurements will be made using one of thetechniques described in Section 4.4

Measurement Techniques.

Calculation of the emission reductions for this

methodology is based on an emission factor

approach. The emissions calculations will be

made according to the following approach:


11/33


Page 10

(1)

(2)

(3)

(4)

Where:

ETB = Total baseline emissions across all manufacturers (tonnes CO2e/yr)

ETi = Total baseline emissions from the ith manufacturer (tonnes CO2e/yr)

N = Number of manufacturers in the project assessment boundary

ETis =Emissions from high-bleed controllers configured for snap-actingcontrol from the ith manufacturer for each identified piece of equipment

ETit =Emissions from high-bleed controllers configured for throttle controlfrom the ith manufacturer for each identified piece of equipment

EFi =Emission factor at 95% lower bound for the ith manufacturer (scfdproduced gas/controller)

GC =

Gas composition of methane in produced gas (mole fraction CH4). If arepresentative sample of GC data is used, this data should bediscounted by a margin-of-error analysis, which can follow the sameprocedure in Appendix B.

CF =Conversion factor for methane volume to mass (0.00001926497tCH4/scf CH4)

GWP = Methane global warming potential (21 tonnes CO2e/tonnes CH4)

ATs,j =Actuation time of snap-acting controllers in basin or facility j(see 3.1.1below)

OiTime

,s=

Percent annual operating time for the ith manufacturers snap-actinghigh-bleed controller (operation hours/total hours per year)

OiTime

,t=

Percent annual operating time for the ith manufacturers throttle actinghigh-bleed controller (operation hours/total hours per year)

3.1.1Calculatingtheactuationrateforsnapacting

controllers

An important concept with certain kinds of

pneumatic controllers called snap-acting

controllers is the actuation time (AT). When

snap-acting controllers are operating and

conducting their intended function, natural gas

is run through the device and is thus not bled to

the atmosphere. This period of functioning,

when no bleeding occurs, is called actuation.

Only in their inactive state are snap-acting

controllers whether high-bleed or low-bleed bleeding gas. Most snap-acting controllers are

inactive, and thus bleeding, the vast majority of

the time (greater than 95%). However, the

project proponent should still take into account

the time the device is active and not bleeding

gas, so this time can be discounted from overall


12/33


Page 11

baseline emissions. For this methodology, the

project proponent will develop a default

actuation rate to apply to the snap-acting

controllers in the project boundary. The default

will be based on gathering data from allcontrollers to be retrofitted and should provide

the most conservative result for each distinct

facility included in the project boundary.

For the production sector a basin-by-basin

approach is necessary because the

actuation rates are directly related to liquid

production volumes and therefore tend not

to vary much across a basin, but may differ

significantly between different basins.

For the transmission sector a facility byfacility (e.g. each compressor station) is

necessary because actuation rates could

vary significantly between facilities

depending on where in the transmission

process the facility is positioned.

The project proponent will collect the following

data to determine the default actuation rate for

snap-acting controllers within a specific facility.

The following data needs only to be collected

for sites that utilize snap-acting controllers, andthis data can be collected at any time:

Port Size. The project proponent will

determine port size for the snap-acting

control valves that are managed by the

controllers to be retrofitted. Port size can

be determined by physical inspection

and should conform to manufacturers

standard port sizes. For the purposes of

determining the most conservative

default actuation rate, the smallest portsize will be used as representative of

the entire retrofit population. If the

smallest size is used, this will be

conservative because the smaller the

port size, the longer the estimated

actuation time would be. Because this

time is deducted from the baseline

emissions, the longer the time, the lower

(and thus more conservative) the

baseline will be.

Differential Pressure. The project

proponent will determine the lowest

differential pressure (DP) across the

snap-acting control valves within the

population to be retrofitted. A common

example of this is the pressure drop

across a dump valve dumping from

separator pressure down to storage tank

pressure. Again, using the lowest

determined DP in the population to be

retrofitted will yield the most

conservative result because lower DPresults in a decreased driving force for

flow through the valves, resulting in a

slower flow rate and decreased liquid

capacity through the control valves;

hence, longer actuation time.

Liquid Capacity: The project proponent

will determine the liquid capacity (LC)

from the valve manufacturer using the

lowest differential pressure (DP) and

smallest port size. The manufacturer ofthe valve should provide the information

(sometimes in a chart called the Liquid

Capacity Chart) in its literature. LC

should be expressed in barrels per day

of water flow rate. By using barrels per

day of water, and NOT correcting for oil

gravity, the calculations will remain

conservative since water is denser than

oil and will result in slightly longer

actuation times. As an example, a

company may have a Kimray valvelocated in an area where the DP is 200

psi and the port size is 1/4 inch.

According to Kimray, in this case, the

liquid capacity would be 650 barrels per

day.


13/33


Page 12

Barrels of Production per Controller

(BPC): The project proponent will

determine the 30-day average combined

(oil or water) production per day for the

well sites with controllers to beretrofitted. The 30-day averages will be

added together to determine the total

combined average volume per day per

controller across all of the controllers to

be retrofitted.

Any controller for which 30 days worth

of production data is not available

should be excluded from the calculation

of the actuation rate.

The percentage of time the actuator isoperating daily (AT) and the snap-acting

controller is NOT bleeding can be expressed

as:

ATs,j= (BPC / LC) where

ATs,j= the default actuation rate in facility jfor

snap-acting (s) controllers

ATs,jis expressed as a percentage or fraction.

Thus, the adjusted baseline bleed rate for a

particular snap-acting controller in facility jwould be the bleed rate * (1-ATs,j). If ATs,jwere

3%, that means that controller is bleeding 97

percent of the time, and the adjusted baseline

emissions would be 97% of the original

baseline calculation.

For pneumatic controllers in the transmission

sector e.g., compressor scrubber dump valves,

the actuation rate should be calculated by

facility e.g., compressor station. The production

rate and therefore the controllers actuation ratecould vary significantly between facilities

depending on where the facility is located in the

transmission network.

3.1.2Thecaseofthrottlepneumaticcontrollers

Calculating actuation times, as described

above, is not necessary in the case of throttle

controllers. With these types of controllers, the

actuator is using a portion of the instrument

supply gas and the remainder is bleeding

through the high-bleed controller pilot, which

provides intermediate valve travel and controls

the process variable (such as level) at a steady

point. Since the actuator is in continuous use,

the atmospheric emissions consist of combined

actuator vent gas and pilot bleed gas that

cannot be delineated from each other. Throttle

configured controllers are used in liquid service

applications as well as gas service applications,such as flow control or pressure control. As with

the snap-acting controllers, the bleed rate can

be determined as described in Section 4.4.

However, no discount for applying an actuation

rate will be required. Thus, the variable ATs,jin

equation (3) does not apply for throttle

controllers.

3.2 ProjectEmissions

An emission factor for estimating project

emissions will be developed from site-specific

measurements taken before each annual GHG

emission reduction verification activity. A

sample of site-specific measurements will be

performed, using procedures outline in

Appendix C, to test the post-project emissions

of 30 different converted controllers. Using the

data from the 30 measurements, a t-test will be

performed, and the emission factor will be

selected as the upper boundary of the 95%

confidence interval. By using the upper boundof the confidence interval, the uncertainty

around sampling is automatically discounted,

thereby providing a conservative result.


14/33


Page 13

If a single technology has been used to retrofit

or replace all devices, a single project emission

factor can be developed. However, if a number

of different low-bleed technologies are being

used for the replacement, then a separate

emission factor must be used for each low-

bleed technology.

The equation to determine project emissions is

as follows:

(9)

(10)

Where:

ETP = Totalprojectemissionsacrossallmanufactures(tonnesCO2e/yr)

EPi = Totalprojectemissionsfortheithmanufacturer(tonnesCO2e/yr)

EFP =Calculatedemissionfactorat95%upperboundfortheithmanufacturer

(scf/day)

Ni = Totalnumberofcontrollersfromtheith

manufacturer

GWP =

Methaneglobalwarmingpotential(21tonnesCO2e/tonneCH4)

GC =

Gascompositionofmethaneinproducedgas(molefractionCH4).Ifa

representativesampleofGCdataisused,thisdatashouldbediscountedby

amarginoferroranalysis,whichcanfollowthesameprocedureinAppendix

B.

CF

= Conversion

factor

for

methane

volume

to

mass

(0.00001926497

tCH4/scf

CH4)

OiTime = Percentannualoperatingtimefortheithidentifiedpieceofequipment(%)

Note: Actuation time is not included in project emissions. This is conservative

because by not calculating the actuation time, project emissions are not discounted.

Thus project emissions would be higher and the emission reductions lower.

3.3 Leakage

Leakage is defined as an increase in emissions

outside the project boundary which is

attributable to the implementation of the project.In cases where leakage occurs, it must be

accounted for and subtracted from the overall

emission reductions for the specific verification

period.

Project proponents should consider two types

of leakage when determining if there is any

leakage attributable to the project. These aresystem leakages and upset leakages,

described as follows:


15/33


Page 14

a) System Leakageis the estimate of

fugitive emissions created elsewhere in

the facility as a result of implementing

the project. As operating practices at

individual companies will vary, eachproject proponent will provide a

conservative method for accounting for

any material system leakage that could

be reasonably expected from the

project.

b) Upset Leakageis the estimate of

leakage due to an increase in upset

conditions as a result of implementing

the project. The project proponent

should consider whether the project is

likely to cause increased malfunctions of

the pneumatic controllers and other

components which would result in

leakage. The project proponent will

provide a conservative method for

accounting for any material upset

leakage that could be reasonably

expected from the project.

Alternatively, if the project proponent deems

that there is neither material system nor upset

leakage, the project proponent may support this

assertion using qualitative and/or quantitative

analysis.

3.4 EmissionReductions

Total emission reductions will be calculated as

the difference between the baseline emissions

and the project emissions.

Emission reductions created by a project will becalculated according to the following equation:

(11)

Where:

ERy = Emissionreductionsduringtheyeary(tonnesCO2e/yr)

ETB,y = Baselineemissionsinyeary(tonnesCO2e/yr)

ETP,y = Projectemissionsinyeary(tonnesCO2e/yr)


16/33


Page 15

4.0 DATA COLLECTION AND MONITORING

This section outlines procedures for data

collection and monitoring. The section

discusses instrument calibration and necessary

parameters for monitoring.

The process for keeping track of the emission

reduction projects parameters includes:

Identifying and logging high-bleed

pneumatic controllers to be converted

Following the instructions for calibrating

measurement instruments Selection of the random sampling for the

project emissions

Oversight of project emissionsmeasurements, such as operating time,

methane concentration, and bleed rates

Recording project emissions data and

calculating project emission factors

The tables at the end of this section list the key

parameters of the emission reductions

calculation, and how each parameter should be

recorded. All information for the project should

be stored electronically and centrally in a single

spreadsheet that is easily accessible for theproject validator and emission reduction

verifiers.

Table4.0:ExampleofPneumaticRetrofitProjectDatabase

4.1 VerificationPeriod

The verification period can be defined at the

discretion of the project proponent, provided it

conforms to ACRs guidelines on verification

periods.

4.2 BaselineEmissionsMeasurement

The emission factors are the foundation of thebaseline emissions calculations. These

baseline emission factors are proscribed by this

methodology and are not directly relevant to

defining the data collection and monitoring

requirements associated with this methodology.

Diligent count and record of the number and

manufacturer of eligible controllers converted to

low-bleed will be required for the methodologys

implementation. The section below elaborates

on necessary data collection for baselineconstruction.


17/33


Page 16

# Parameter Procedure

1 Number of controllers Project manager should keep a catalog of all of the controllersintended to be replaced by the project, including manufacturer andcontroller type. This database will be used for the selection of theproject group in subsequent GHG emission reductions verifications.

2 Location Record of location of the pneumatic controller within the operationsinfrastructure is critical in order to account for the operational status ofthe equipment.

3 Low-bleed pneumaticcontrollers

Each converted pneumatic controller should be given a uniqueidentifying number that shows it to be a part of the project assessmentboundary. For each potential retrofit, the project proponent shouldconfirm that the retrofit has taken place.

The project proponent is responsible for keeping track of whichretrofits have already been used in sampling and use controller andbasin data to select subsequent sample groups.

Given that the major equipment to which the pneumatic controllers areassociated has operations which are monitored routinely; operatinghours for each low-bleed pneumatic controller will be recorded.

Each controller should include a description of whether it is continuousbleed or interrupted bleed.

4.3 ProjectEmissions

Measurement

4.3.1Selectionofverificationsamplinggroup

Quantification of project emissions will require

direct measurement of emissions. Direct

measurement of emissions will be taken from a

random and representative sample of 30 post-

retrofit pneumatic controllers, and a project

emission factor will be calculated for each

verification period.

4.3.2Identification

of

verification

sampling

group

Once a controller has been selected for the

verification sampling group, it should be noted

in the data base that that the controller hasbeen selected. After the emissions test has

been performed, the controller should have a

durable identifier (e.g. documented GPS

coordinates, unique identifier code, stamped

metal tag) so that it is easily identifiable and will

be found by the verifier or project manager in

subsequent data quality QC spot-checks, GHG

emission reduction spot-checks, or verification

site visits. Finally, pneumatic controllers in the

sample verification group should be excludedfrom the random selection for the subsequent

verification, and then re-included in the pool

thereafter.


18/33


Page 17

4.3.3Dataanalysisandemissionfactorderivation

foremissionfactorsamplinggroup

The 30 data points will be used to determine

the 95% confidence interval for the average

emissions.5 The value at the upper bound of the

95% confidence interval should be selected as

a conservative measure.

4.3.4Outliersintheverificationsamplegroup

All the emissions from all controllers selected

for the verification sample group should still be

included in the sample mean calculation of the

project emission factor. If a control is found to

be bleeding gas at an abnormally high rate, the

project proponent should determine if thereading is a typographical error or a

measurement inaccuracy and take the

necessary step to correct this reading.

However, if the abnormally high reading is

accurate it should still be included in the project

emissions calculation. No outliers will be

removed unless they are inaccurate

measurements and cannot be rectified.

4.3.5Ensuringconsistencyinbaselinesample

measurements

In order to ensure that emissions

measurements are in accordance with normal

operating parameters, pneumatic gas supply

pressure for controllers within the snap-acting

sample set must fall within manufacturers

specifications. Controllers for which pneumatic

gas supply pressure readings fall outside

manufacturers specifications should be

excluded from the sample set.

5Duetothehighdegreeofcontrolonboththe

calibrationoftheinstrumentsandtherandom,

representativenatureofthesample,a95%upperbound

isappropriateforthedevelopmentoftheproject

emissionfactor.A95%intervalisinlinewithIPCCand

UNFCCCCDMbestpracticeswhenarandomand

representativesampleistaken.

4.4 MEASUREMENTTECHNIQUES

Measurement of emissions will be performed

using one of the following techniques:

4.4.1A

high

volume

sampling

tool

This can include a Hi-Flow Sampler or

comparable instrument which meets the

instrumentation requirements as outlined below

to the satisfaction of the verifier. Due to the low

rate of gas flow of individual pneumatic

controllers, a Hi-Flow Sampler is uniquely

suited to accurately quantify flow rates from

low-emissions (several hundred scfd) sources.

The Hi-Flow Sampler operates by pulling

emissions away from the leaking equipmentinto a centralized conduit for measurement,

automatically correcting for the hydrocarbons in

the ambient air that would otherwise contribute

to the total leak rate measured.

The Hi-Flow Sampler utilizes dual function

catalytic oxidation/ thermal conductivity sensors

to measure the rate of gas flow from the leak

and ensure the accuracy of the measurement

test. The Hi-Flow Sampler measures only

combustible gases (methane, propane, butane,etc) and not inert gases like carbon dioxide and

nitrogen. Therefore when correcting for gas

composition in the afore-mentioned equations

3, 4 and 10, the gas composition correction

factor (GC) should be calculated as the volume

of methane as a proportion of all combustible

gases (methane, propane, butane etc).

The following procedures should be followed in

using the Hi-Flow Sampler to measure

emissions: Calibration on instruments should be

checked, at a minimum, according to the

procedures and frequency outlined by the

manufacturers of the Hi-Flow Sampler.


19/33


Page 18

o For gas concentration,

instrument should be calibrated

using gas that has been

provided by a certified vendor

(Approved within 2%).o For catalytic oxidation, calibrate

for 2.5% methane.

o For thermal conductivity,

calibrate for 99% methane.

Measurements need only be taken

once. The test should endure for a

minimum of 5 minutes to ensure that the

gas emissions are measured at a steady

state.

The leak flow rate of methane is calculated as

follows:

Where:

FCH4,i =The leak flow rate of methane for leak ifrom the leaking component

(scfm)Fsampler,i = The sample flow rate of methane for leak i(scfm)Csample,i = The concentration of methane in the sample flow from leak i(volumepercent)Cback,i = The concentration of methane in the background near the component(volume percent)

4.4.2Inlineturbinemeters

Another option for determining leak rates is the

use of an in-line turbine meter. In this case, a

meter is installed into the system as opposed to

an external sniffing device like the Hi-Flow

Sampler. The process works by shutting supply

pressure off to the controller and isolating it

from service. In most cases, there is a regulator

line feeding the controller. The meter is installed

downstream of the regulator output port by

using inch tubing and fittings compatible with

the regulator and the input port of the flow

meter. Another inch length of tubing is

installed from put the output port of the flowmeter to the input port of the control device

being tested. This ensures the flow meter is

installed upstream of the device.

Before the retrofit from high-bleed to low-bleed,

the system pressure is turned on, slowly

allowing the controller to stabilize to its

operating condition. The flow meter will register

and record the flow of supply gas, which is

being used by the controller. Once the test is

completed, the inlet and outlet shutoff valvesare closed again, isolating the controller from

supply pressure. After the retrofit program the

same test procedure is performed on a different

sample set. The difference between the pre-

retrofit and post-retrofit measurements indicates

the bleed savings opportunity. Once the testing

is completed, the flow meter is uninstalled and

the system is returned to its original

configuration.

All in-line turbine meters shall be calibratedaccording to the manufacturers specifications

prior to measurement.

4.4.3Calibratedbagging

Calibrated bag measurements use anti-static

bags of known volume (e.g. 0.085 m3or 0.227

m3) with a neck shaped for easy sealing around


20/33


Page 19

the vent. Measurement is made by timing the

bag expansion to full capacity while also

employing a technique to completely capture

the leak while the inflation is being timed. The

measurement is repeated on the same leaksource numerous times (at least 7, typically 7 to

10 times) in order to ensure a representative

average for the fill times (outliers or problem

times should be omitted and the tests rerun

until a representative average rate is

established). The temperature of the gas is

measured to allow correction of volume to

standard conditions. Additionally, gas

composition is measured to verify the

proportion of methane in the vented gas, sincein some cases air may also be vented, resulting

in a mixture of natural gas and air. Calibrated

bags allow for reliable measurement of leak

flow rates of more than 250 m3/h. The leak flow

rate of methane is calculated as follows:

FCH4,i=VbagxwsampleCH4,ix3600/taver,iWhere:

FCH4,i =The leak flow rate of methane for leak ifrom the leaking component(m/h)

Vbag = Volume of calibrated bag used for measurement (m)

wsampleCH4,i =The concentration of methane in the sample flow from leak i(volumepercent)

taver,i = Average bag fill time for leak i(seconds)

4.5 DataandAnalysisforVerification

Table 4.5 below details the data to be collectedor calculated.


21/33


Page 20

Table4.5:DatatobeCollectedorCalculated

Parameter Description Data Unit

Calculated [c],Measured [m],Reference [r],

Operatingrecords [o]

Measurement

frequency

ETB Total annualemissions acrossall manufacturersin the baseline

tCO2e [c] Calculated once In cases where only one brandETB= ETi

ETi Total annualbaselineemissions fromthe from the i

th

manufacturer

tCO2e [c] Calculated once

N Number ofmanufacturers inthe projectassessmentboundary

# [c] Calculated once

TNi Total number ofcontrollers fromthe i

th

manufactureracross the project

# [m] Logged at thebeginning of theproject. Changesand updates madeduring eachverification period.

The total populations of each pneumatic controllers to be reinventoried by the project propbrand, and date of conversion recorded electronically and ce

EFi Baseline emissionfactor at 95%lower bound for i

th

manufacturer

scfd [c] Calculated once


22/33


Page 21

6Foss,M.M.,InterstateNaturalGasQualitySpecificationsandInterchangeability,UniversityofTexasBureauofEconomicGeolog




Measurementfrequency

GC Gas compositionof methane

% [m] Frequencydetermined by welloperator

The methane composition of tcalculation element in estimatithe GHG emission reduction pconstituent of natural gas, andproduced in the U.S. is found tbe determined by gas chromatechnology. Depending on theSection 4.4), GC may represethe total gas composition or ofconcentration testing may occoperator according to the charparticular well. In this case, theanalysis will be used in baselin

sampling approach is used, it error analysis, which can followProject proponent may also waStandards and Technology (N

GWPCH4 Global warmingpotential of CH4

21 [r] N/A 21 was determined in the 1996Report to be the Global WarmThird Annual Assessment Reppotential to be 23. Common indictated that 21 should be usemethane in the future.

OiTime Operating hours ofthe productionequipment relatingto the pneumaticcontrollers

% [o] [c] Data to be takenfrom operatingrecords ofcompany

Hours of operation used to cacontrollers on operating equipm


23/33


Page 22




Measurementfrequency

ETP Total projectemissions acrossall manufactures

tCO2e [c] Each validationperiod

Note that in the event that onlyETP= EPi

EPi Total annualproject emissionsfor the i

th

manufacturer

tCO2e [c] Each validationperiod

EFP Emission factor forthe project takenat the upperbound of the 95%confidence interval

scfd [m] [c] Calculated foreach verificationperiod

ETP Total annualemissions for theproject

tCO2e [m] [c] Calculated foreach verificationperiod

BPC Barrels of waterproduced percontroller

Averagebarrels perday/controller

[c] Once Can be measured at any time.period of time snap-acting con

LC Liquid capacity Barrels perday ofwater flowrate

[r] Once Can be measured at any time.period of time snap-acting con

DP Differentialpressure

Psi [r] Once Can be measured at any time.snap-acting controllers.

Port size Port size ofcontroller

Inches [r] Once Can be measured at any time.snap-acting controllers.


24/33


Page 23

5.0 EMISSIONS OWNERSHIP AND QUALITY

5.1 Statementof

Direct

Emissions

The project proponent should attest that all

emission reductions occur on the property

owned and/or controlled by the project

proponent and that none of the emissions

claimed by the project are indirect emissions.

5.2 Permanence

Project proponents should demonstrate that the

emissions reduced by the project are

permanently reduced.

5.3 Title

Project proponents should provide evidence

that they have title or contractual rights to the

emission reductions claimed in the project

document prior to ACR registration and that no

other entity has a conflicting claim over the

emission reduction claimed in the in the project

document.

5.4 Communityand

Environmental

Impacts

It is unlikely that this particular type of project

the simple replacement of controllers at various

and remote oil-production locations wouldhave any significant environmental or social

impacts. The retrofitting of controllers are not

capital-intensive projects requiring any kind of

environmental impact statement, and no new

infrastructure is put in place. Retrofitting

pneumatic controllers only replaces a tiny part

of a large infrastructure already on the ground,

with the environmental impacts having been

incurred long before the retrofit project.

If there are any significant impacts frompneumatic retrofits, however, project

proponents should take into account ACRs

Community and Environmental Impacts criteria

when applying this methodology to specific

projects. These criteria require analysis of any

adverse environmental or social impacts on the

communities near the project boundary. If there

are any adverse environmental problems that

may result from the project itself, project

proponents should conduct stakeholder

meetings with community groups or other

appropriate NGOs to ensure that any concerns

associated with the project are addressed.


25/33


Page 24

6.0 QA/QC

As part of the Pneumatic Conversion Protocol,the number of continuous high-bleed pneumatic

controllers to be converted in the project will be

monitored by the project proponent, with all

pneumatic controllers that are converted being

identified and numbered. As a QA/QC check on

the retrofit of pneumatic controllers, purchasing

invoice and shipping receipt information or

other documentation from the low-/no-bleed

pneumatic controller equipment supplier for the

pneumatic controller conversions should be

available. More details on the specific QA/QCactivities required by the Pneumatic Conversion

Methodology are summarized in Table 7.1 in

the following section.

In addition, in order to ensure that controllers

are retrofitted correctly and to minimize any

chance of future upset conditions resulting

from the retrofit, project proponents should

review the manufacturers recommended retrofit

procedures and the manufacturers

recommended post-retrofit QA/QC checklist.


26/33


Page 25

7.0 UNCERTAINTIES

The emission reduction calculations in this

methodology are designed to understand

possible emissions, and to minimize the

possibility of overestimation and over crediting

of GHG emission reductions, due to various

uncertainties, primarily associated with

emission factors and representative sampling

populations.

Nonetheless, there are a number of

uncertainties related to estimates of methane

emission reductions from continuous high-bleed

pneumatic controller conversion projects. Some

of these uncertainties are more easily quantified

than others. These sources and relative

magnitude of uncertainties (and changes

thereof) should be explicitly addressed and

discussed by the project proponent as part of

the GHG emissions calculation and reporting

process, and described in the project document

as part of the GHG emissions calculation and

reporting process.

These potential sources of uncertainty include

but are not limited to: emission factors,

methane gas composition, operating time,

actuation time and conversion rates. Each of

these potential sources of uncertainty, and the

associated QA/QC program elements designedto minimize them is summarized in Table 7-1.

Other factors as stated above (e.g. equipment

overhaul frequency, instrument settings) have a

bearing on uncertainty. Overall uncertainty can

be assessed by using the uncertainties of each

element in a calculation.

The accuracy and precision of measurement

equipment, such as the high flow meter and gas

composition analyzer, are readily quantified and

the uncertainties associated with eachmeasurement are considered to be quite low.

The extrapolation of gas composition

measurements to the total bleed rate of

methane gas can introduce some uncertainty,

depending upon the variability of the methane

content of the gas (geographically and

temporally) and the frequency of samples

taken.

Operating outages in the baseline and the

project activity will be recorded, if and when

they occur, though the contribution of this

variation to the overall uncertainty is not

expected to be large and conservative

estimates of durations will be made.

Table7.1:UncertaintiesintheReportedData

Data ParameterUncertainty

Level of DataQA/QC Procedures Risk Mitigation

Emission Factors(EFi,Band EFP)

Low/Medium EFi,B-Field instruments calibrated before use, statisticalanalyses at 95% confidence interval, and projectemission factor at lower bound

EFP-Field instruments calibrated before use, statisticalanalyses at 95% confidence interval, and projectemission factor at upper bound


27/33


Page 26

Data ParameterUncertainty

Level of DataQA/QC Procedures Risk Mitigation

Operating Parameters Low/Medium Supply Pressure Pneumatic gas supply pressure of

controllers sampled for the purposes of determining the

baseline and project emissions will be recorded at the

time of sampling and must be within manufacturer

recommended specifications

Maintenance Conditions After retrofits have been

completed, overhauls to correct upset conditions should

be documented. Additionally, estimates of emissions

resulting from the upset condition should also be

recorded under Leakage as described in Section 3.3

Instrument

Parameters

Low/Medium Actuation Rate Sampling must only be conducted

when the controller is in a steady state and not actuating.Additionally, sample measurements need to be taken fora time period in excess of 5 minutes to ensurestabilization of the gas emissions and avoid any spikes

Liquid levels and Proportional Band Controllers mustbe installed and configured to operate withinmanufacturers specifications e.g. within manufacturersspecified bands and levels

Number of High andLow-Bleed Pneumatic

Controllers (Ni,)

Low Each replaced controller should be tagged with a durableunique identifier (e.g. stamped metal, GPS coordinates

etc) capable of lasting several years and inspected infield, with count confirmed with equipment suppliersales/installation records

Methane GasComposition (GC)

Low Multiple measurements sources; historicaltract/field/reservoir/basin/compressor station data forvalidation

Operating Time(OiTime)

Low Project entity maintains detailed and continuousinformation on the operation of the equipment

Actuation time forsnap-acting controllers

(AT)

Low AT is calculated with data that should be available to alloil companies, including port size of the controllers to be

retrofitted, the differential pressure and barrels ofproduction. All are basic data parameters that wouldeither be available centrally or by observing in the field.Liquid capacity is provided by the manufacturer of thevalves.


28/33


Page 27

APPENDIX A

This annex presents a summary of the data, summarized in Table A.1, used to prove that there are

statistically significant differences associated with manufacturer design differences in continuous high-

bleed pneumatic controllers. This observation is also confirmed in the vendor literature for low-/no-

bleed pneumatic controllers.7 Based on that, as part of the statistical analysis of the baseline sampling

test data, stratified data sets (by manufacturer) were also evaluated.

Table A.18: High-Bleed Pneumatic Controller Emissions Data Summary

7Wellmark,MizerPilotControl,ContinuousBleedControllersLostGas.www.wellmarkco.com

8TableA1summarizesthehighbleedemissionsdatapointstakefromthemeasurementof35Cemco(18)andInvalco(17)controllers.

Datawasprovidedbytwo(2)oilandgasproductioncompaniesandthelow/nobleedpneumaticcontrollermanufacturer,withsampling

measurementstakenbytheirfieldtechniciansfortwodifferentoilandgascompaniesrepresenting8differentproductionfieldsin4

states.

Manufacturer Model Baseline Bleed Rate, scfd Data Source

Cemco 6900 367.0 Producer 3

Wellmark/Cemco 6900 432.0 Producer 3



Cemco -- 463.2 Producer 3

Cemco 6900 @ 25 psi 576 Producer 3

Cemco 6900 @ 35 psi 696 Producer 3




Cemco Snap-acting 401.3 Producer 1

Cemco Throttle 386.0 Producer 1

Cemco Throttle w/ snap head 441.9 Producer 1






Invalco Throttling 518.1 Producer 1

Invalco Snap-acting 274.3 Producer 1


29/33


Page 28

Statistically Significant Difference between Average Emissions for Cemco and Invalco

The sample average for Cemco controllers was

511 scfd and the sample average for Invalco

controllers was 629 scfd. Given that the

emissions from the different manufacturers are

different, it becomes imperative to assess if

there is a statically significant difference

between the emissions from Cemco controllers

and Invalco controllers.

If there is a statistically significant

difference in the sample means then it

indicates that the differences between the

sample means is not due to chance, but

rather due to real differences in controllers,

e.g. design. The implication of this is that

baseline sampling must be carried out for

each controller manufacturer

independently (i.e. 30 samples of Cemco

controllers and 30 samples of Invalco

controllers).

If however, there is not a significant

difference between sample means, then it

indicates that the difference between the

sample means is likely due to chance and

not as a result of real differences. Therefore

the Cemco and Invalco controllers can be

considered as one population from a

baseline estimation point of view (i.e. 30

samples of controllers).

Analysis: To assess statistical significance a

two sample ttest for independent samples was

conducted. Assuming unequal variances,

tcalculated = 1.76 with a p-value = .091. This p-

value is less than the significance level of .10

which leads to the conclusion that there is a

statistically significant difference between the

Cemco and Invalco average emissions.

Manufacturer Model Baseline Bleed Rate, scfd Data Source

Invalco CTS-215 655.0 EPA

Invalco CTU-215 1052.2 Producer 2

Invalco CTU-415 597.6 Producer 3Invalco CTU-415 559.2 Producer 3













30/33


Page 29

In other words, the difference in sample means

is notdue to chance but is the result of real

differences in the controllers. Therefore the

standard baseline emission factors for high-

bleed controls cannot be directly appliedwithout manufacturer specific measurement

data and the project proponent must sample

different manufacturers controllers as separate

populations.

Estimating the Average Emissions for

Cemco and Invalco

Table A.2 summarizes the results of the

confidence interval calculations for the

individual manufacturers. Margins of error and

confidence intervals were calculated using the

standard method assuming a normal population

distribution and an unknown population

variance. See Appendix B for a detailed

explanation of confidence interval calculation

methods. The resulting values are summarized

in the table below.

Table A.2: 95% Confidence intervals for mean emissions of Cemco and Invalco controllers

Mean Value (scfd) Stdev Margin of Error Lower Bound Upper Bound

Cemco 511 147 73 437 584

Invalco 629 238 122 507 751


31/33


Page 30

APPENDIX B

This appendix defines the procedure for

calculating an interval estimate for thepopulation mean, . Because it is rarely

possible to take a census of a population of

items, we must rely on the results of a sample

data set to estimate the mean of a population.

The interval estimate is called a confidence

interval. The typical form of any confidence

interval is to add and subtract a margin of error

from a sample statistic.

When estimating a population mean, the

sample statistic used is the sample mean, .

The margin of error of any interval estimate is

dependent on three factors: the level of

confidence chosen by the analyst, the variation

within the population, and the size of the

sample used for estimation.

The formula for calculating a confidence interval

for is:

(1)

where is the margin of error, tis the

statistic used to guarantee the confidence of the

estimate, sis the sample standard deviation,

and nis the sample size. The margin of error of

any interval estimate is dependent on these

three values.

The order in which one proceeds in the

estimation process is the following:

Choose a confidence level. Typically, the

confidence level used is 95%. By choosing

a higher confidence one would be

increasing the margin of error, all things

being equal. By choosing a lower

confidence one would decrease the margin

of error, all things being equal. In other

words, greater confidence leads to wider

intervals, lesser confidence leads to more

narrow intervals. Determine the appropriate sample size.

The size of the sample should be

determined by the precision level

desired (expressed as the margin of

error) and confidence level desired and

will be dependent on the amount of

variation in the population. Depending

on how these three factors change, the

appropriate sample size could go up or

down. For example, we can use thesample standard deviation calculated

from the Invalco data in App A (s= 238

scfd) as an estimate of the population

variation. If we choose a confidence

level of 95%, and choose a margin of

error of 100 scfd, the appropriate

sample size would be n= 22. If one

were to use a larger sample size, the

margin of error would be less than 100

scfd, as long as the population variation

has not increased.

Collect the sample and calculate the

sample mean, and the sample

standard deviation, s.

Determine the value of t, the statistic

matched to the desired confidence level.

For example, if using a sample size of

30 and a desired confidence level of

95%, the tstatistic value is 2.045. The

value of the tstatistic is based on a

parameter called degrees of freedomwhere df= n 1. The tstatistic value

can be found on any standard Students

ttable of critical values.

Calculate the confidence interval limits

using formula (1).


32/33


Page 31

Example calculations: Refer back to the sample

data provided for the baseline emissions of the

Cemco and Invalco controllers shown in

Appendix A. A confidence level of 95% was

used in the calculations. There are many off-

the-shelf statistical software packages that will

perform these routine calculations with the

appropriate sample data input on a

spreadsheet.

Manufacturer Cemco Invalco

Sample size, n 18 17

Sample mean, 510.5 628.9

Sample Std Dev, s 147.2 237.9

tstatistic 2.110 2.120

Margin of Error 73.2 122.3

95% Confidence Interval (437 to 584) (507 to 751)


33/33


APPENDIX C

This appendix defines the procedure forsampling emissions during the estimation of the

baseline and project emissions. The project

proponent will take direct measurements, using

measurement techniques outlined in Section

4.4 above, from a sample of a minimum of 30

pneumatic controllers that are correctly installed

and operating under manufacturers

recommended guidelines. In addition to

emissions measurements the project proponent

will also collect the following information at the

time of measurement:

Unique controller identifier

Supply pressure across the controller

Port size (snap-actuating only)

In addition to the data collection, the project

proponent will, at the time of sampling, do the

following:

Visual observation of correct installationof the pneumatic controller

Observed successful operation of the

controller, which might require physically

triggering the controller

Additionally, the project proponent may choose

to increase the size of the sample

measurements to a number greater than 30. By

doing this the project proponent will get a more

accurate estimate of the population average

emissions and a smaller margin of error at the95% confidence level, which could mean a

higher lower bound for the baseline emissions

estimate and a lower upper bound for the

project emissions estimate and therefore

potentially a greater net overall emission

reduction value. This is illustrated in Appendix B

and by the following table.

Sample size (n) Minimum size: 30 Greater than 30 (e.g. 50)

Degrees of Freedom (n-1) 29 49

t statistic at 95% confidence level 2.045 2.012

Margin of error formula

Margin of error 0.373 S30 0.284 S50

S30is the standard deviation of a random set of 30 emissions measurements and S50is the standard

deviation of a random set of 50 emissions measurements. Typically, S50> S30due to the larger samplesize and therefore the margin of error for the 50 random sample will be less than that for the random

sample of 30 measurements.

ACR Methodology for the Conversion of High-Bleed Pneumatic Controllers in Oil and Natural Gas...

Documents

Transcript of ACR Methodology for the Conversion of High-Bleed Pneumatic Controllers in Oil and Natural Gas...